Scientists Just Proved AI Therapy Chatbots Are Breaking Every Rule in the Book

Millions of people are turning to ChatGPT and other AI chatbots for therapy-style advice. New research from Brown University says that's a really bad idea.

The study found that even when these AI systems are specifically told to act like trained therapists, they consistently break the core ethical rules that every real therapist has to follow. We're talking about things like maintaining boundaries, protecting patient confidentiality, and knowing when to refer someone to emergency services.

Think of it this way: imagine going to a doctor who looks like a doctor, talks like a doctor, but never actually went to medical school and doesn't follow any medical rules. That's essentially what's happening when people pour their hearts out to AI chatbots.

The timing is especially concerning because AI therapy apps are booming right now. They're cheaper than real therapy (often free), available 24/7, and don't have a waiting list. For many people, especially younger users, chatting with AI feels less intimidating than talking to an actual person.

But the Brown University researchers are warning that this convenience comes with real risks. When someone is in crisis, an AI that doesn't know the ethical playbook could give dangerous advice or miss warning signs that a trained professional would catch immediately. As reported by ScienceDaily.

Source: ScienceDaily